- i studied Electronics & Computer Engineering at Thapar Institute of Engineering & Technology, Patiala, Punjab, 🇮🇳
- i’m currently working as a Full-Stack/Applied AI Engineer as well as a Consultant at Cubastion Consulting, Gurgaon, India. apart from this, i'm also quite familiar with the ML/DL domain.
- i’m looking to collaborate and participate in hackathons and other tech competitions.
- ask me about cars and music.
- how to reach me: my Linktree
- fun fact: i'm a hobbyist graphic designer, video editor, writer and an aspiring actor (since the first time someone asked me what i wanted to do when i grew up).
- click on the headings below to view more about me!
[1] average fresher, starting out with baby steps
a crash course in the building blocks of applied AI engineering:
- Company Database APIs: Python, Flask, PostgreSQL
- faceMatching: computer vision with PyTorch
- Milvus Querying: embeddings, vector search and similarity retrieval
- Cubastion HR Chatbot: my first practical RAG application
- Joined the GEN AI - FUSO team.
- Co-developed the TIC Chatbot for Mitsubishi FUSO - an enterprise-grade AI assistant with a vast scope.
- Learned what changes when an AI application has to work beyond a notebook.
built projects around semantic resume matching, my first AI agent, and PDF processing/conversion - moving from individual ML experiments toward systems that could actually do things.
started R&D on a Smart HR Chatbot with a custom agentic architecture, capable of handling HR queries, referrals, leave workflows, manager feedback and employee benefits through conversation.
Python · LangChain · Streamlit · Azure OpenAI · Pinecone · MySQL · Linux
[2] stood on my feet and instantly signed up for the marathon
experiments turned into systems.
today, i work across AI engineering, document intelligence, full-stack development and cloud infrastructure, building the future of enterprise innovation in the form of cutting-edge custom solutions from architecture through deployment.
i've worked on systems where the interesting problem isn't simply “how do we call an LLM?” - it's everything around it.
- reworked a large-scale service-document translation pipeline, taking it from poor initial translation quality to substantially higher accuracy while bringing 100+ page processing down to minutes through concurrency, caching and pipeline redesign.
- helped architect the backend of an enterprise chatbot with 30+ APIs, bringing together RAG, semantic retrieval, access control, analytics and both vector and relational data stores.
- built an image/document translation workflow connecting Adobe Illustrator scripting, a packaged Node.js application and LLM-powered translation on Azure - including working with proprietary publishing formats.
- co-leading development of an Agentic AI platform combining visual workflow orchestration, MCP-based tool calling and multi-agent execution.
my work has increasingly moved toward the layers that make AI applications useful in the real world:
LLMs · RAG · Agentic AI · MCP · FastAPI · Node.js · PostgreSQL · Vector Databases · Azure · Docker · CI/CD
basically, the job isn't just building the system - it's figuring out what should be built in the first place, working with business stakeholders and client leadership to turn messy requirements into things that can actually ship.
